Environmental science and technology researchers alike are ardently studying the voracious energy demands and carbon emissions of machine learning and artificial intelligence, searching for "green AI." Efforts like Digital Cleanup Day urge the average user to tidy up their overflowing inboxes to reduce the pressure on hot, panting cloud servers. But "Green AI" surely will not be green enough. And when data storage gets too tight, most of us respond by loosening the belt on iCloud, OneDrive, or Google Docs and proceeding to gobble data at the same rate as before.
In the enterprise, there is often an assumption that more data, better data, more powerful computers, more intelligent computers will automatically provide life-saving, earth-shattering, business-empowering insights. And they can. Or they can just take up space, waste money, waste energy, and generally make a mess. Respondents to a recent InformationWeek study told us that "data analytics and data-driven initiatives" were the main reason that their data volumes were increasing at leaps and bounds, and 90% acknowledged that environmental sustainability is or should play a role in getting a handle on data management.
So we're devoting an entire week of coverage to the topic. If AI, machine learning, data analytics, and supercomputing could potentially take a toll on the planet, how can you run them most responsibly and efficiently? How can CIOs and CDOs balance sustainability goals and digital transformation?
Here are some of the articles that are part of this series:
A New Generation and the Future of Sustainable Computing
By Carrie Pallardy, Contributing Reporter
The Gen Z generation has grown up with both powerful technology and a keen awareness of environmental impact. How will their perspectives as the new data scientists and stakeholders shape the future of sustainable computing?
How DPUs, IPUs, and CXL Can Improve Data Center Power Efficiency
By Sal Salamone, Managing Editor, Network Computing
New technologies like DPUs, IPUs, and CXLs offload switching and networking tasks from server CPUs. They have the potential to significantly improve the data center power efficiency and sustainability.
Read the rest of the article on InformationWeek.
